High prevalence of methicillin-resistant Staphylococcus aureus carriage among infants at the Children’s Hospital, Accra, Ghana

Introduction: Infants are at risk of Staphylococcus aureus ( S. aureus ) colonization and infection. The aim of this study was to investigate S. aureus and methicillin-resistant S. aureus (MRSA) colonization among infants, including the prevalence, predictors of colonization, and antibiogram. Methodology: The study was cross-sectional, and involved infants aged less than one year recruited at the Princess Marie Louise Children’s Hospital in Accra, Ghana. Sociodemographic and clinical data of the participants were gathered with a structured questionnaire. Nasal swabs were also obtained from them and bacteriologically cultured. S. aureus was confirmed with the coagulase test, and MRSA was confirmed by polymerase chain reaction (PCR) of the mecA gene. Antimicrobial susceptibility testing of S. aureus was done using the Kirby-Bauer method. Results: The carriage prevalence of S. aureus and MRSA were 34.9% (45/129) and 17.10% (22/129), respectively. Colonization with coagulase-negative Staphylococci (CoNS) was protective of both S. aureus (OR = 0.008; p < 0.001) and MRSA (OR = 0.052; p = 0.005) carriage. Maintenance of good hand hygiene prevented S. aureus carriage (OR = 0.16; p < 0.001). S. aureus resistance to antibiotics decreased across penicillin (96%), trimethoprim-sulfamethoxazole (61%), tetracycline (61%), erythromycin (39%), gentamicin (39%), fusidic acid (26%), rifampicin (17%), clindamycin (7%), and linezolid (0%); 68.8% S. aureus were multidrug resistant. Conclusions: S. aureus and MRSA prevalence were high among the infants. Colonization with CoNS and good hand hygiene maintenance were predictive of MRSA and methicillin-sensitive S. aureus (MSSA) colonization, respectively.


Introduction
Staphylococcus aureus (S. aureus) preferentially colonizes the moist squamous epithelium of the anterior nares, and less frequently, other anatomical sites [1]. The clinical significance of its colonization is reflected in it being an antecedent to subsequent infections such as endocarditis, pneumonia, septicaemia, and meningitis [2,3]. Moreover, the treatment of its infections has been complicated by the occurrence and spread of its multidrug resistant strains, particularly, those referred to as methicillin-resistant S. aureus (MRSA), which are refractory to all antibiotics of the beta-lactam class -a trait that their methicillin-sensitive contemporaries called methicillin-sensitive S. aureus (MSSA) lack [4]. The molecular basis of this methicillin resistance is the encoding of a novel penicillin-binding protein with lower affinity to penicillin -PBP2a -by the acquired cassette-borne mecA gene and its variants thereof, such as the recently discovered mecC gene [4]. Similar to some other pathogens such as Clostridium difficile, biofilm formation occurs extensively in some S. aureus strains, and this contributes to the relatively high levels of antibiotic resistance in MRSA [5,6,7]. Besides the mec genes that mediate multidrug resistance in S. aureus, several virulence genes of the pathogen have been identified, including the Panton-Valentine leukocidin (pvl) gene, whose gene product causes lysis of polymorphonuclear white blood cells and tissue necrosis [7,8]. Although the pvl gene typically occurs in community-associated MRSA, it is frequently encountered among S. aureus strains circulating in Africa, including methicillin-sensitive ones [9][10][11][12].
Globally, between 1970 and 2011, the proportion of S. aureus infections that were attributed to MRSA in reports had increased from 2% to 80% [12][13][14][15]. MRSA is one of the major causes of septicaemia among neonates, and is associated with a 20-40% mortality rate, regardless of appropriate treatment [16,17]. It is also one of the common pathogens associated with hospitalization, morbidity, and mortality among children below five years of age in developing countries. In addition, its infections result in extended hospital stays and increased healthcare costs [18,19]. According to a report from the United States, over 94,000 new invasive MRSA infections occur annually, resulting in more than 18,000 deaths [20]. In another report, over 250 hospitalization days and almost 30,000 MRSA bloodstream infections were reported in 31 European countries, with an associated death toll of 5,503 [18].
Infants are frequently associated with healthcare settings, owing to their need for postnatal care and their young immune systems that predispose them to diseases requiring medical attention [21,22]. As a matter of concern, some studies have demonstrated that exposure to healthcare settings predispose to S. aureus and MRSA carriage [23,24]. Therefore, it is not surprising that infants have been identified as a risk group for S. aureus carriage, and this makes them potential reservoirs for dissemination of the pathogen to other individuals [25]. The increasing proportion of S. aureus strains that are MRSA, carriage of the pathogen being a precursor of its infections, as well as the possession of a relatively weaker immune system by infants collectively make this risk group more vulnerable to increased case fatality resulting from MRSA infections [3,[10][11][12][13][14][15]18,19]. Several MRSA outbreaks have been recorded in Ghana since 2012 [26], underscoring the public health threat the pathogen poses in the country, and concurrently giving credence to its surveillance among risk populations with the purpose of assisting with public health interventions. The risk groups that have been the focus of the limited MRSA carriage studies in the country are sickle cell disease patients [27], HIV-infected persons [24,28], and children [29,30], and these studies have reported carriage prevalence of 0-15%. Infants have not been well studied in relation to carriage of MRSA, despite the potentially high clinical significance of the pathogen among them. To fill this important knowledge gap, this study investigated S. aureus and MRSA colonization among infants, focusing on prevalence and risk factors for carriage, as well as the pvl gene carriage of MRSA and the antibiogram of colonizing S. aureus.

Study site, design, and sampling
This study was carried out using the Princess Marie Louise Children's Hospital (PML) in Accra as the study site. Accra is the capital city of Ghana, and is home to approximately two million individuals, as well as eight government or quasi-governmental hospitals (http://www.statsghana.gov.gh/). Besides the Korle Bu Teaching Hospital, which includes numerous specialties and provides specialized paediatric care, PML is the only other major public hospital in Accra that has pediatric care as its focus. This was a crosssectional study involving 129 healthy infants visiting the outpatient clinic of the hospital, either for postnatal care or by virtue of accompanying their guardians for their routine medical examinations. Sampling of the participants was done between January and July 2017, and their parents/guardians gave consent for their participation. The inclusion criteria for participation in the study were: being an infant below one year of age, being in a steady state, and being an outpatient. The exclusion criteria were: being on antimicrobials two weeks prior to sampling and presence of comorbidities (with the exception of asthma, which was found in two of the study participants). Data on potential S. aureus and MRSA carriage determinants were gathered from the infants using a standard questionnaire.
A qualified paediatrician rotated a sterile swab stick in both anterior nares of each of the infants. The resultant swab specimens were maintained on ice in uniquely-labeled 1 mL skim milk-tryptone-glucoseglycerin (STGG) medium-containing vials, and after four hours following collection, they were transported to the Department of Medical Microbiology, University of Ghana Medical School, for laboratory processing. Each of the swab specimens were vortexed for approximately two minutes, and thereafter refrigerated at -80 °C, until needed.
The molecular methods used in the investigation of the isolates are as follows: The manufacturer's instructions of the Zymo Research extraction kit (Zymo Research Corp., Irvine, USA) were used as a guide to extract genomic DNA from overnight lysogenic broth cultures of a S. aureus positive control strain (ATCC 25923) and the MRSA isolates. In order to check the quality of DNA, 5 µL of each extracted DNA was mixed with 2 µL of bromophenol blue gel loading buffer and ran on a 1.2% agarose gel; the resultant bands were visualized by UV illumination. The extracted genomic DNA was used as template for the PCR amplification of each of the three genes -mecA, nucA, and pvl -in a total reaction volume of 50 mL, made up of the genomic DNA (60 ng/µL final concentration), deoxyribonucleoside triphosphates (dNTPs) (200 µM final concentration), MgCl 2 (2 mM final concentration), Taq polymerase (1.25 U/µL final concentration), primers (0.2 µM final concentration), and PCR water; RNase-free water was used as the negative control. The primer sequences used in the amplification of the mecA, nucA, and pvl genes were those described by Sajith Khan et al. [32], Brakstad et al. [33], and Deurenberg et al. [34], respectively. Separation of the amplicons was done using 1.2% agarose gel electrophoresis, followed by visualization with the aid of UV illumination.

Data Analysis
The data were analyzed using STATA 14 (Strata Corp, College Station, TX, USA). Descriptive statistics were used to summarize data on the resistance of S. aureus to the tested antimicrobials. A combination of independent samples Chi square tests, point biserial correlations, and binary logistic regression were used to identify determinants of S. aureus and MRSA colonization. The significance of each determinant was evaluated using its p value, odds ratio, and confidence interval; p values whose magnitudes were below 0.05 were deemed significant.

Ethical Approval
This study was approved by the Ethical and Protocol Review Committee of the College of Health Sciences, University of Ghana, with protocol identification number "CHS-Et/M.3 -P 4.4/2016-2017".

Demographic, household, and clinical features of the infants
The one hundred and twenty-nine (129) infants recruited as participants in this study were predominantly male (58.1%, n = 75), and had a mean age of 18.29 weeks. Furthermore, majority of them were assisted by their caregivers to maintain good hand hygiene (55.8%, n = 72), resided in compound houses (86%, n = 111), with an average of 4.19 persons living in these households, but did not have any member of their households employed in healthcare (100%, n =  Table 1.

129). A summary of the demographic and household characteristics of the infants is presented in
The majority of the participants had not been hospitalized (67.4%, n = 87), and had no history of surgery (98.4%, n = 127), asthma (98.4%, n = 127), or pneumonia (98.4%, n = 127). The clinical features of the infants are presented in Table 2.

Carriage of S. aureus and MRSA, and associated determinants
In the current study, 17.1% (n = 22) of the infants carried MSSA exclusively, and 17.8% (n = 23) carried MRSA exclusively, making the total carriage prevalence of S. aureus 34.9% (n = 45); none of the participants had MRSA-MSSA co-colonization. The carriage prevalence of pvl-positive MRSA was 16.28 (n = 21); hence the proportion of MRSA isolates that were pvl-positive was 84% (21/25). Conversely, coagulasenegative Staphylococci (CoNS) were present in 40.3% (n = 52) of the infants, and none of them concurrently carried S. aureus.
The logistic regression analysis revealed that of the variables evaluated as risk factors for colonization, colonization with CoNS was protective of both S. aureus and MRSA colonizations. Moreover, maintenance of good hand hygiene was also found to be protective of S. aureus colonization, specifically, S. aureus isolates that are methicillin-sensitive, but not methicilin-resistant. The results of the risk factor analysis are presented in Table 3.

Antibiogram of the S. aureus isolates
The highest rate of antimicrobial resistance was recorded against penicillin, whereas no resistance was recorded against linezolid. Moreover, the proportion of S. aureus isolates that were multidrug resistant (those that were resistant to three or more classes of antimicrobials) [24] was 68.8%. The prevalence of antimicrobial resistance among the S. aureus isolates are presented in Figure 1.

Discussion
The current study investigated the epidemiology of S. aureus and MRSA nasal carriage among infants. It is the first of its kind in the country, and among the few such studies in this risk group globally. The carriage prevalence of S. aureus was found to be 34.9%, and the respective prevalence for MSSA and MRSA were 17.1% and 17.8% respectively. The S. aureus and MRSA carriage prevalence seem higher than those reported in a study conducted among children in the country -22.1% and 2%, respectively [30]. This difference might be accounted for by the fact that Eibach et al.'s study [30] included children aged up to   more than ten years, a wider age range than that in this study which focused on infants (< 1 yrs). Indeed, age has been identified as an important determinant of S. aureus carriage [30,35]. Another noteworthy observation that makes the disparity in prevalence between the current study and that of Eibach et al. [30] quite paradoxical is that the participants sampled in that study were admitted in the hospital with a variety of microbial infections and hospital admission is a known risk factor for S. aureus carriage [23,27,36,37]. The S. aureus prevalence recorded in the current study is lower than the 49% recorded by Donkor & Nartey [29] in their study involving children under five years of age, and this could reflect a temporal decline in S. aureus carriage, given that sampling of the participants for the two studies were done about a decade apart. This may especially be the case, as nasal carriage prevalence of 10-23.4% has been reported among healthy individuals in more recent studies [24,27,36]. A high proportion (84.1%) of the MRSA isolated in the current study harboured the pvl gene. Although the MSSA isolates were not screened to determine their carriage of the gene due to limited resources, it is recognized that this high proportion of pvl-positive MRSA feeds into the widely accepted hallmark of S. aureus strains circulating in Africa -that they frequently harbour pvl [9][10][11][12]. Moreover, carriage of the pvl gene is often associated with CA-MRSA [38,39].
In the current study, colonization with CoNS was found to be protective of both S. aureus and MRSA colonization, indicating that absence of colonization with these organisms predisposes to carriage of S. aureus and MRSA. Donkor et al. [24] made a similar finding, albeit among HIV-infected children. The finding is further supported by the reports of Iwase et al. [40], Olson et al. [41], and Paharik et al. [42]. Iwase et al. [40] and Olson et al. [41] demonstrated that Staphylococcus epidermis, a part of CoNS, prevents S. aureus colonization. Paharik et al. [42] provided the mechanism underlying this antagonism -CoNS prevents S. aureus colonization by inhibiting agr quorum sensing in the latter using its autoinducing peptide. This suggests that the autoinducing peptide, as well as other CoNS-produced peptides that are cidal to S. aureus, could be further investigated for their usefulness in S. aureus decolonization strategies.
Maintenance of good hand hygiene was also found to be protective of S. aureus colonization, specifically, S. aureus isolates that are methicillin-sensitive, but not methicilin-resistant -this distinction regarding the protective value of good hand hygiene against bacterial colonization is an intriguing finding. Overall, it is at odds with the study conducted by Donkor et al. [24] cited above, which reported that regular hand washing with soap and a good hand hygiene practice increased MRSA colonization by more than six folds, with a possible link to the use of triclosan-containing soap. It is noted that in the current study the investigation of good hand hygiene practice did not make a distinction between the use of soaps for hand washing and the application of hand sanitizers and other preparations made for the promotion of hand hygiene. That notwithstanding, this finding fits into the long-held custom of promoting good hygiene practices as a means of controlling the spread of infections of several pathogens [43][44][45][46][47][48]. It is therefore imperative to intensify the promotion of good hand hygiene practice, particularly, among populations who are at a higher risk of S. aureus carriage.
The high rates of resistance that were recorded against penicillin, trimethoprim-sulfamethoxazole, tetracycline, fusidic acid, erythromycin, and gentamycin are consistent with those reported in other studies in the country [24,27,36], and can be attributed to their increased use over the years. It is interesting that our study revealed a high prevalence of trimethoprimsulfamethoxazole resistance among S. aureus (which are probably community-associated), contrary to the observation that trimethoprim-sulfamethoxazole resistance tends to be more often linked with hospitalassociated S. aureus [49,50]. This observation could be due to a number of reasons; the widespread routine use of trimethoprim-sulfamethoxazole in hospital and community settings has generally promoted resistance to the antibiotic among hospital and communityassociated S. aureus isolates circulating in the country. The observation may also reflect a changing epidemiology of S. aureus with respect to hospital and community settings. Further studies using molecular methods are required to further understand this. The low resistance rate (0%) recorded against linezolid was expected, given that the drug has a low coverage, and access to it is restricted. The low rate also suggests that the drug could retain its place as part of the mainstays of anti-MRSA therapy. However, given the high proportion (68.8%) of multidrug resistant S. aureus in the current study, it would not be surprising if subsequent studies in the country begin reporting increasing resistance rates against the antimicrobial. It is therefore important to intensify antimicrobial stewardship programmes in the country.
This study was potentially limited by a few factors. First, molecular typing of the isolates was not done due to the high costs of the technique. Also, screening for the pvl gene was done for only the MRSA isolates, and hence the proportion of MSSA isolates that harboured the gene is unknown. Moreover, the study does not distinguish between intermittent and persistent S. aureus carriers, as for every participant, anterior nasal sampling was done only once.

Conclusions
It is concluded that the prevalence of S. aureus and MRSA were high among the infants, who may serve as reservoirs of multidrug resistant S. aureus. Colonization with coagulase-negative Staphylococci is a determinant of both S. aureus and MRSA colonization, while maintenance of good hand hygiene is a determinant of MSSA colonization. The autoinducing peptide produced by coagulase-negative Staphylococci could be further investigated for its usefulness in S. aureus decolonization strategies.